Can you lose your accent? What research on adult learners shows

What does it mean to lose your accent?

Before asking whether you can lose your accent, remember that everyone speaks with one. When people talk about losing an accent, they usually mean swapping features linked to one language or region for features of another. That could mean changing a few vowels, learning a different r, or sounding close to a model you have chosen. It can't literally mean speaking with no pronunciation pattern at all.

These outcomes overlap, but they are not the same thing. In their 1995 study, Munro and Derwing compared listeners' ratings with how accurately those listeners could write down what they heard. A stronger perceived accent did not necessarily mean people understood less. So it makes sense to measure clarity directly, rather than assuming every noticeable difference gets in the way.

Pick a goal you can observe. Be easier to understand in project meetings is more useful than never sound foreign. You can still work on imitating an accent for personal or professional reasons. Keeping these goals apart helps you decide whether a session should focus on contrasts that change meaning, on rhythm, or on fine regional detail.

Is there an age after which you can't change your accent?

The critical-period question asks whether biological development makes native-like language harder or impossible to reach after a certain age. For pronunciation, it helps to separate a general age-related trend from a fixed deadline for one person. Saying that everyone loses the ability to learn new sounds on a particular birthday is a much stronger claim than the evidence that early and late learners often end up differently.

Bongaerts, van Summeren, Planken and Schils (1997) studied late learners, including a hand-picked group of very successful Dutch learners of English. Some of them received accent ratings within the range of the native speakers in the study. That challenges the idea that a late start makes native-like pronunciation impossible. Because the group was selected, though, it doesn't tell us how common that result is among learners in general.

The practical lesson is modest. Age alone is no reason to give up on pronunciation. It also doesn't mean anyone can reach the same result with enough effort. Your learning history, how much English you hear, and what exactly is being measured all matter. A short rated recording can't show how someone will sound when they are tired, emotional, or in an unfamiliar conversation.

What does Flege's Speech Learning Model say about adult accents?

Flege and Bohn's revised Speech Learning Model, published in 2021, proposes that the ability to learn speech sounds stays available throughout life. It explains differences between learners partly by the sound categories they already have, how different a new sound seems from a familiar one, and how much good-quality input they get. It is a theory, not a training programme with guaranteed results.

One simple way to use that idea is to look at a contrast you have been treating as a single sound. If ship and sheep seem the same to you, you may need to hear the vowel difference clearly before you can keep the words apart when you speak. Following the spelling or repeating a rough version may leave you with the same single category.

For /ɪ/ and /iː/, use a model recording and compare tongue position as well as length. For /w/ and /v/, round your lips without touching your teeth for /w/, and press your upper teeth against your lower lip for /v/. These exercises are practical ideas based on the learning problem. They are not an experimental test of the model.

Does pronunciation training work? What Derwing and Munro found

Derwing, Munro and Wiebe (1998) followed three groups through a twelve-week course. One group was taught individual sounds, one was taught broader speaking habits and prosody, and one had no specific pronunciation teaching. Both pronunciation groups improved on rated sentences. Only the broader group improved in comprehensibility and fluency when telling a story spontaneously.

That result is a good argument for including connected speaking in your practice. It doesn't prove that sound drills are useless, that one programme suits everyone, or that twelve weeks will give you the same outcome. The gap between prepared sentences and storytelling matters most here. A skill that works while you read may need more practice before it holds up in real-time speech.

Look at a study's year and design together. An older, influential study can answer a narrow question well and still leave questions open about other languages, audiences or modern tools. The sources at the end link to the original publications, so you can tell the reported findings apart from the practice suggestions in this guide.

How adults can work on their accent without chasing total accent loss

Start with a real communication task, such as a short work update. Record it, find two unclear words, and decide whether the problem is a sound, the stress or the timing. Practise that feature in a simpler phrase, then give the update again. You are looking for a change you can hear and actually use.

• Listen to a short model containing your chosen feature.
• Repeat slowly, changing one mouth movement or stress pattern.
• Record a prepared sentence and compare it with the model.
• Use the feature in a new sentence without a script.
• Ask a listener which words or details they understood.

Then switch tasks. Explain where to leave a document, describe a place you have visited, or compare three options. Use your target sound wherever it fits naturally. If you can only say it correctly in the original sentence, give shorter answers and build up again.

The pronunciation improvement guide covers the full loop. For a day-by-day schedule, use the 30-day accent plan. Treat a month as a chance to check your habits and progress, not a deadline for turning into a different speaker.

How to measure accent progress without chasing perfection

Save a baseline reading and a short unscripted recording. Every few weeks, repeat the reading and record a new answer on a similar topic. Keep the microphone position and background noise about the same. Look at your individual targets and at how well listeners understand you before you judge the overall impression.

Flalingo's accent test estimates accent strength from 0 to 100, where lower means closer to its native reference. Its band and native-similarity percentage can't prove that you now sound like a native speaker, and the test doesn't name your variety. Use it as one point of comparison next to your recordings.

The pronunciation test gives word and phoneme scores, plus feedback on fluency, completeness and intonation, measured against a General American reference. Here a higher score means a closer match to the model, the opposite direction from accent strength. Neither score tells you everything about a real conversation. If listeners catch your words more reliably, that is useful evidence, even if your overall score hasn't changed much.